[Annual stipend of £21,935 (as of 2024-25 fiscal year, subject to revision) for 3.5 years, plus £5000 research expense funds. Tuition fees + stipend available for Home/EU and International students. Enhanced stipend possible via SPADS CDT for eligible UK applicants.]

This PhD studentship at The University of Edinburgh focuses on advancing machine learning methods for dynamical systems in sensor signal processing. Dynamical system models are foundational in control, signal processing, and sensor fusion, enabling applications such as multi-object detection and tracking, robotic SLAM, and calibration of autonomous networked sensors. The project aims to address challenges in learning hierarchical, time-varying multi-dimensional state space models for dynamic objects and phenomena, especially when dealing with noisy measurements, complex backgrounds, and calibration errors.

Recent advances in machine learning have shown that leveraging data and model size can mitigate model inaccuracies, leading to significant performance gains in inference and decision-making. However, learning from data in dynamical system models to jointly address epistemic and aleatoric uncertainties remains challenging due to factors like noisy data, inhomogeneous sampling, model complexity, and intractable posterior inference. The research will explore solutions to these issues, with opportunities to steer the direction towards engineering problems such as radar detection in complex backgrounds, birth and trajectory models for improved detection and tracking, and semi-supervised or unsupervised training of sensor data classifiers.

Supervised by Dr M Uney (expert in signal processing, machine learning, probabilistic models, and Bayesian computation) and Prof Mike Davies (Jeffrey Collins Chair in Signal and Image Processing, specializing in machine imaging and sensor fusion), the successful candidate will join a vibrant research environment at the School of Engineering. The position is open to UK/EU and international applicants, offering an annual stipend of £21,935 (as of 2024-25 fiscal year, subject to revision) for 3.5 years, plus £5000 research expense funds. Tuition fees are covered for Home/EU and International students. UK nationals and eligible applicants may align the studentship with the Sensing, Processing and AI for Defence and Security Centre for Doctoral Training (SPADS CDT), potentially benefiting from an enhanced stipend (subject to CDT approval and security clearance).

Eligibility requires a minimum of a 2:1 Master’s Degree in engineering, computer science, statistics, physics, or a related discipline, and compliance with the University’s English language requirements. Preferable qualifications include a 1st class undergraduate or Master’s degree in relevant fields and experience in scholarly writing. Applications are also welcomed from self-funded students or those seeking scholarships from the University of Edinburgh or elsewhere.

To apply, submit your application online via the University of Edinburgh portal, and indicate your interest in the SPADS CDT in your Personal Statement if applicable. For further information, visit the project webpage or contact the supervisors directly.